-
Notifications
You must be signed in to change notification settings - Fork 20
Mount extensions directory to persist generated data #71
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
| - ./config:/mediawiki/config | ||
| - ./images:/mediawiki/images | ||
| - sitemap:/mediawiki/sitemap | ||
| - ./extensions-data:/mediawiki/extensions |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think it's confusing to call this /mediawiki/extensions when the mount is called extensions-data and contains more than just extension data - it also contains extensions code!
|
CanastaWiki/CanastaBase#35 brings up a good point, that the persistence of extension data is not currently possible. Unfortunately, this is a very tricky issue to solve. MediaWiki does not do a good job of giving extensions the ability to separate ephemeral data from persistent data, even though this is fundamental to the mantra of Docker. Code for extensions should be considered ephemeral. My guess is that this folder is going to be a combination of ephemeral extension execution code and the persistent data written by certain extensions. Please correct me if I am wrong. By combining extensions into this extensions-data mount, we are risking sysadmins messing with extensions-data by accident. Is there an alternative place to put this so sysadmins won't accidentally start putting stuff into this folder? Can we put this on ice for a few days while the Christmas holidays wrap up, so that we can get some more comments from folks? |
|
@cicalese, this might be a good opportunity for an architectural consultation by you. :) |
|
Noting that this would not resolve the theoretical issue of skins needing a persistent directory and that @jeffw16's concerns are valid. However, the good news is that none of that matters. I think this PR is actually not needed and that CanastaWiki/CanastaBase#44 solved both issues I reported in CanastaWiki/CanastaBase#35 (comment) by fixing both the wrong timing and the wrong location. Although the title of that PR implies it only fixed the location of the commands, it did also fix the path structure. Now that the locations are correct, the |
|
That's great to hear! I guess I'll mark the original bug as closed. And yes, I'm aware that "persistent directories" is still not handled for skins - I don't think this is a problem, though, because I can't imagine a skin ever needing to store its own data. |
|
To clarify, I believe that (in theory) persistent skins directories should work now too. I'll close this and the bug report and we can reopen if needed. |
Oh! Good to know. |
Fixes CanastaWiki/CanastaBase#35